KSP Spaceplane C/G and design help

[overdrive148]

So, I've been playing a good amount of Kerbal Space Program lately, and while I have rockets down to a fine art, spaceplanes are a major achilles heel. And being a pilot and all, that's a challenge I can't turn down.

Here's my first serious attempt:

I originally had two intakes and two engines per side, but they were not producing enough air/thrust to get me airborne and the climb rate was anemic. Performance at altitude was mediocre. I'm using "Rapier" engines, which perform like normal engines but can also use oxidizer/liquid fuel to add quite a bit of thrust (and work in space of course).

On the climb to altitude, I end up burning a good chunk of fuel of course. I'm noticing that the SAS (stability augmentation system, basically heading hold AP) is having to really push pitch limits in cruise to stay level. I toggled the SAS off and immediately flipped over on my back. I realized the problem was most likely CG changes due to fuel burn and started moving fuel around the tanks to the nose to see if I could balance it out on the fly.

After that didn't work, I shifted the canards to pitch only, took two of the 4 roll elevons (2nd inboard) and turned those to pitch only and also engaged the reaction wheels (electricity => control) and STILL couldn't keep the nose down.

Here's the layout before takeoff, full fuel.

Blue is center of lift, black and yellow are center of mass/gravity, pink is the thrust line. I didn't have enough oxidizer onboard for any useful amount of burn so I added a couple drop tanks in the space between the fuselage and the nacelles. The C/G dropped straight down, so it's behind and below the C/L. Flies decently without SAS on after takeoff. Adding more mass to the nose and bringing the C/G before the C/L any farther than this makes the nose drop - this was the balance I struck after flight testing and adjusting.

Here's the fuel drained out of the all of the tanks, including drop tanks:

The C/G moves waaaaay far back. The amount of fuel or weight I'd need add to balance out the CG again would make it -very- nose heavy at takeoff. There's no totals for gross weight but from how it flies it feels pretty close to it. The best thing I can think of is to add a couple empty tanks up front and as the flight progresses, move the fuel from the back to the front tanks and lock them to use them as ballast or emergency fuel.

Is it a bad design or is it salvageable?

 

[Henning]

Your CG needs to go way forward, it's costing you too much work and drag to keep stable.

The further forward the CG the more corrective arm you get.

 

[EdFred]

I stay in the VAB. When I've messed around in the SPH I always run into the CG CoL issues as well. The problem is the engines are too heavy for their size. They really need to be longer and integrated more into a nacelle. The other issue is you can't put fuel in the wings. You should go into your part cfgs and make it so you can put fuel in the wings. It's only 4 lines of code.

 

[Henning]

See if you can get your CG around 35% and your center of drag around 65% back from the nose, and follow the Area Rule as you go back and see what you get. I see a major wing redesign in your future.

 

[Henning]

Look at the X-29 forward sweep wing plan, I think that is the direction you need to look, maybe an asymmetrical diamond shaped flying wing.

 

[EdFred]

Look at the X-29 forward sweep wing plan, I think that is the direction you need to look., maybe a diamond shaped flying wing.

You obviously haven't gotten involved in KSP.

 

[Henning]

You obviously haven't gotten involved in KSP.

No, it doesn't work on an iPad.

 

[EdFred]

No, it doesn't work on an iPad.

Yeah, it's not just "move the wing" in KSP. The parts' masses are unbalanced. As soon as you put the engines on, everything goes way back. And the physics engine doesn't allow for them to be forward - at least not easily.

 

[Henning]

Yeah, it's not just "move the wing" in KSP. The parts' masses are unbalanced. As soon as you put the engines on, everything goes way back. And the physics engine doesn't allow for them to be forward - at least not easily.

There is always a way to build it regardless what it is, you just have to keep looking in other directions until you figure it out. With a forward swept wing in a forward cruciform/diamond shape you can put the engines on the wing tips and move them as far forward as you want.

Think flying Trident fork forward type proportionality, the engines could even lead.

 

[overdrive148]

I stay in the VAB. When I've messed around in the SPH I always run into the CG CoL issues as well. The problem is the engines are too heavy for their size. They really need to be longer and integrated more into a nacelle. The other issue is you can't put fuel in the wings. You should go into your part cfgs and make it so you can put fuel in the wings. It's only 4 lines of code.

Agreed on the engines, they are pretty heavy. I didn't know you could change the code for the wings to have fuel in them, that would go a long way to balancing them without adding more weight underneath them.

Your CG needs to go way forward, it's costing you too much work and drag to keep stable.

The further forward the CG the more corrective arm you get.

True, but keeping the CL forward enough to make up for it is turning out to be pretty hard.

See if you can get your CG around 35% and your center of drag around 65% back from the nose, and follow the Area Rule as you go back and see what you get. I see a major wing redesign in your future.

I tried standard wings and diamond shaped and neither of them got me off the runway Not enough "meat" to the wing I don't think.

Look at the X-29 forward sweep wing plan, I think that is the direction you need to look, maybe an asymmetrical diamond shaped flying wing.

I've tried the forward sweep on other non-space planes. I'll try throwing a pair on here but with the way this is laid out already, it seem unlikely to work.

Yeah, it's not just "move the wing" in KSP. The parts' masses are unbalanced. As soon as you put the engines on, everything goes way back. And the physics engine doesn't allow for them to be forward - at least not easily.

Engines on vs engines off, empty tanks

 

[EdFred]

Just put this in the .cfg file

RESOURCE
{
 name = LiquidFuel
 amount = n
 maxAmount = n
}

make n whatever number you feel appropriate for that part.

If you need to use oxidizer just add those exact lines of code again, and change LiquidFuel to Oxidizer.

Also remember the rapier engines burn 9:11 LF:O ratio when out of the atmosphere, or whenever the air intake drops to 0.

 

[EdFred]

There is always a way to build it regardless what it is,

Not in KSP.

 

[Henning]

Agreed on the engines, they are pretty heavy. I didn't know you could change the code for the wings to have fuel in them, that would go a long way to balancing them without adding more weight underneath them.


True, but keeping the CL forward enough to make up for it is turning out to be pretty hard.


I tried standard wings and diamond shaped and neither of them got me off the runway Not enough "meat" to the wing I don't think.


I've tried the forward sweep on other non-space planes. I'll try throwing a pair on here but with the way this is laid out already, it seem unlikely to work.


Engines on vs engines off, empty tanks

Turn the engines around, flip the control surfaces and cockpit around and try it just like that.

 

[overdrive148]

Turn the engines around, flip the control surfaces and cockpit around and try it just like that.

Interesting idea, I'll try it when I get back to the house.

 

[Henning]

Interesting idea, I'll try it when I get back to the house.

If you ever get a chance to watch a stingray in nature you realize what a magnificently efficient and adaptable shape that it is.

 

[overdrive148]

If you ever get a chance to watch a stingray in nature you realize what a magnificently efficient and adaptable shape that it is.

Not home quite yet but am I reversing the gear in relation to flight or leaving them exactly as they are? Trying to think of anything else.

 

[Henning]

Not home quite yet but am I reversing the gear in relation to flight or leaving them exactly as they are? Trying to think of anything else.

The craft is going to fly wide end forward if that is your question. The point of this craft is to be able to fly off the ground and into space right?

 

[overdrive148]

The craft is going to fly wide end forward if that is your question. The point of this craft is to be able to fly off the ground and into space right?

Yep.

So here we have it:

Changed landing gear direction, engines, intakes, control surfaces. Everything else is in the same positions as you suggested.

Left is CG with fuel, right is CG without fuel. The capsule on the tail end is ejected for equal weight to the original and is there because it's not easy to remove it and have another capsule as primary.

Flight testing in progress, hard hats on.

 

[overdrive148]

Not cropped because effort.

Basically impossible to get in the air no matter the landing gear distance from the C/G (close to it or farther) including the same configuration and distance in the original. SAS on makes it get about half way down the runway before it slips. SAS off means the kerbals die much quicker.

I tried to correct it in the video but obviously it didn't work. Full deflection including using reaction wheels and 3 of 4 surfaces on each wing dedicated to pitch didn't get the nose up at all. Looks to be an unsuccessful model.

 

[overdrive148]

And your other suggestion of flip the wings forward only:

Flies like you expect it to with that far aft C/G

 

[Henning]

It looks like you may need to go Pod Racer format and put the engines out in front of a Flying V or Y planform since it looks like all the weight is at the back end of the engines. You need a geometry that puts the CG of the engines as close to the, but forward of the CL as possible.

Can you mount them to vector thrust?

 

[overdrive148]

It looks like you may need to go Pod Racer format and put the engines out in front of a Flying V or Y planform since it looks like all the weight is at the back end of the engines. You need a geometry that puts the CG of the engines as close to the, but forward of the CL as possible.

Can you mount them to vector thrust?

Most of the air breathing engines have small amounts of thrust vectoring, and can be toggled on and off. Will see if I can design something in a Y or V. Control surfaces are going to be hard to place in advance I think.

 

[Henning]

Most of the air breathing engines have small amounts of thrust vectoring, and can be toggled on and off. Will see if I can design something in a Y or V. Control surfaces are going to be hard to place in advance I think.

Can you pivot the entire engine mount for vectoring?

 

[overdrive148]

Not with stock parts. I think there are mods out there that will let you but I haven't come across them personally

 

[John221us]

More cowbell

 

[Henning]

Not with stock parts. I think there are mods out there that will let you but I haven't come across them personally

If you could mount them "tilt rotor" style with even 5° available it would help, but even a bit of nozzle tilt if the nozzle is around 35% length should have a useful benefit.

 

[overdrive148]

Heading out of the house again but I have tweaked it enough to get it to rotate but it won't leave the ground until it runs off the end of the runway. Turning ability is weak, probably due to the required small elevator surfaces to keep the C/L about that far from the CG. Any bigger and it drags the C/L back hard.

Thinking I might need more wing for that kind of fuel capacity but trying to add it and make it look and act properly is a beast. The front canards are slaved to pitch only for rotation ability and turning - without them it is basically impossible to pull the nose up due to the non-rigid attachment of the pod carrying the elevators (wiggles the pod instead of the whole thing).

 

[Henning]

Heading out of the house again but I have tweaked it enough to get it to rotate but it won't leave the ground until it runs off the end of the runway. Turning ability is weak, probably due to the required small elevator surfaces to keep the C/L about that far from the CG. Any bigger and it drags the C/L back hard.

Thinking I might need more wing for that kind of fuel capacity but trying to add it and make it look and act properly is a beast. The front canards are slaved to pitch only for rotation ability and turning - without them it is basically impossible to pull the nose up due to the non-rigid attachment of the pod carrying the elevators (wiggles the pod instead of the whole thing).

Hmmm, can you create a collapsible biwing arrangement? How about high lift devices like flaps and slats? Are these engines very weak or something?

 

[Henning]

BTW, my vision had the wing as a forward wept V it's the engines on the tips, and a lifting body creating a Y if it needed the extra lift.

 

[overdrive148]

Hmmm, can you create a collapsible biwing arrangement? Are these engines very weak or something?

Collapsible probably not, biwing probably yes. No flaps or slats either in the vanilla version of the game.

Craft is 38.7 tons gross, engines provide 175 kN of force each. Each engine weighs about 1.2 tons.

BTW, my vision had the wing as a forward wept V it's the engines on the tips, and a lifting body creating a Y if it needed the extra lift.

This is the podracer style attempt ofc, testing for some viability.

 

[Henning]

Collapsible probably not, biwing probably yes.

Craft is 38.7 tons gross, engines provide 175 kN of force each. Each engine weighs about 1.2 tons alone.



This is the podracer style attempt ofc, testing for some viability.

So low power to weight for the job. See my post above and edit prior about high lift devices.

 

[Henning]

Then there is also a conjoined wing design that eliminates the tip vortices and this drag.

 

[mkosmo]

Are you using FAR? Stock aerodynamics suck hard... Drag is a product of mass alone. FAR fixes quite a bit (including surface area and drag in general) but it's still not perfect. It also introduces mach effects, so be ready for that fun. FAR will also let you display IAS/EAS/Mach instead of ground speed.

 

[warthog1984]

Your wings are too big, your canards oversized, and your plane too stubby.

There's a reason delta wing planes always look like lawn darts- try lengthening the fuselage, shrinking the wings, and dropping the canards.

 

[Henning]

Your wings are too big, your canards oversized, and your plane too stubby.

There's a reason delta wing planes always look like lawn darts- try lengthening the fuselage, shrinking the wings, and dropping the canards.

Exactly, lifting body.

 

[overdrive148]

Then there is also a conjoined wing design that eliminates the tip vortices and this drag.

Your wings are too big, your canards oversized, and your plane too stubby.

There's a reason delta wing planes always look like lawn darts- try lengthening the fuselage, shrinking the wings, and dropping the canards.

Exactly, lifting body.

Problem is, KSP is far from an "accurate" aerodynamic model. No part of the ship other than the wings produces lift Also vortices aren't modeled AFAIK.

Shrinking the wings down makes me think that I'll need a lot of thrust to get/stay in the air, and then we're dealing with more of a winged rocket instead of a space plane. Hm.

Also canards basically come in one size and shape unless I add a 'wing' and an elevator, which combined add more lift and displace the C/L more than just the canard alone.

Are you using FAR? Stock aerodynamics suck hard... Drag is a product of mass alone. FAR fixes quite a bit (including surface area and drag in general) but it's still not perfect. It also introduces mach effects, so be ready for that fun. FAR will also let you display IAS/EAS/Mach instead of ground speed.

I'll look into that. I used B9(?) Aerodynamics pack in the past but it was completely off. The testing here is off of 100% stock KSP.

 

[mkosmo]

B9 just includes parts, no change to the aerodynamics modelling. If FAR is too much to chew, the author has a much simpler version called NEAR that may be worth your time. NEAR skips things like mach compressability that FAR accounts for.

Sent from my SAMSUNG-SGH-I337 using Tapatalk

 

[overdrive148]

B9 just includes parts, no change to the aerodynamics modelling. If FAR is too much to chew, the author has a much simpler version called NEAR that may be worth your time. NEAR skips things like mach compressability that FAR accounts for.

Sent from my SAMSUNG-SGH-I337 using Tapatalk

Hm, maybe it wasn't B9. I forget, but it definitely changed the way most of my aircraft flew. I'll look into that, thanks!

 

[mkosmo]

I should probably warn you that neither will likely impress a pilot, but it's at least far more predictable. The other outstanding issue is the thrust to weight of the engines and such, but other mods address that. Also, the patch for wet wings above will help. I'm tempted to write a module manager patch to do some of this.

 

[mkosmo]

Exactly, lifting body.

But this is ksp we're talking about lol. Ignore most of reality and add more boosters and struts if you want it to work better.

If you've never played it, it's probably the single most amusing physics sandbox out there. Worth every penny in my opinion. Pick it up!